JP5399497B2 - MMSEMIMO decoder using QR decomposition - Google Patents

Description

  The present disclosure relates generally to MIMO-OFDM communication systems, and more particularly to techniques for removing bias in MMSE detection based on QR decomposition.

A multiple-input multiple-output (MIMO) communication system employs multiple (N _t ) transmit antennas and multiple (N _r ) receive antennas for data transmission. N _t transmit antennas and N _r receive antennas and MIMO channel formed by the be decomposed into N _S independent _{channels, N S ≦ min {N t} , N r} is satisfied. Each of the N _S independent channels is also referred to as a spatial subchannel of the MIMO channel and corresponds to one dimension. MIMO systems improve performance (eg, transmission capacity) over the performance of single-input single-output (SISO) communication systems when additional dimensionality generated by multiple transmit and receive antennas is utilized. Increase).

  Broadband MIMO systems generally undergo frequency selective fading, which means various amounts of attenuation over the system bandwidth. This frequency selective fading causes intersymbol interference (ISI), a phenomenon in which each symbol in the received signal acts as a distortion to subsequent symbols in the received signal. This distortion degrades performance by affecting the ability to correctly detect received symbols. Thus, ISI is a non-negligible noise component that can have a significant impact on the overall signal-to-noise interference ratio (SNR) of systems designed to operate at high SNR levels, such as MIMO systems. In such a system, channel equalization can be used at the receiver to eliminate ISI. However, the computational complexity required to perform equalization is generally significant or prohibitive for most applications.

Orthogonal frequency division multiplexing (OFDM) can be used to eliminate ISI without using computationally intensive equalization. An OFDM system effectively partitions the system bandwidth into several (N _F ) frequency subchannels, sometimes referred to as subbands or frequency bins. Each frequency subchannel is associated with a respective subcarrier frequency that can modulate data. The frequency subchannel of an OFDM system depends on the characteristics of the propagation path between the transmit antenna and the receive antenna (eg, multipath profile), and frequency selective fading (ie, different amounts of different frequency subchannels). (Attenuation). In OFDM, as known in the art, ISI due to frequency selective fading can be eliminated by repeating portions of each OFDM symbol (ie, adding a cyclic prefix to each OFDM symbol). . Thus, a MIMO system can advantageously employ OFDM to eliminate ISI.

  In order to increase the transmission data rate and spectral efficiency of the system, spatial multiplexing can be utilized at the transmitter to transmit various independent data streams over multiple spatial subchannels. The detection accuracy at the receiver may be significantly reduced due to strong multiple access interference (interference of various data streams transmitted from multiple antennas). Further, the spatial and frequency subchannels of a MIMO-OFDM system may be subject to different channel conditions (eg, different fading and multipath effects) and may achieve different SNRs. In addition, channel conditions can vary over time.

  In order to mitigate the effects of multiple access interference, noise and fading, minimum mean square error (MMSE) channel equalization is generally applied at the receiver. However, the estimated signal obtained from the MMSE algorithm contains a bias representing self-noise that reduces detection accuracy. It is well known in the art that the MMSE technique can be simplified if QR decomposition of the expanded channel matrix (hereinafter abbreviated as QRMMSE detection) is used instead of direct inversion of the channel matrix. However, in this particular case, direct bias removal is computationally complex. In addition, the calculation of noise variance required for outer channel decoding may be difficult.

  Therefore, there is a need in the art for a simpler and more efficient technique proposed in this disclosure for removing bias from signals obtained after MMSE equalization. The proposed scheme also provides substantially improved detection accuracy compared to bias MMSE detection.

  Some embodiments of the present disclosure provide a method for performing bias removal from a received signal obtained after minimum mean square error (MMSE) equalization in a wireless communication system. The method generally includes permuting an expansion matrix of channel estimates of multiple spatial streams to generate a permutation expansion matrix of channel estimates of multiple spatial streams and unitary matrices of the multiple spatial streams and an upper matrix. The received signal of the first spatial stream to perform QR decomposition of the permutation augmentation matrix to generate a triangular matrix and to generate a bias filtered output of the first spatial stream of the spatial streams Are rotated with a corresponding unitary matrix and the biased filtered output is subjected to a diagonal element of the corresponding upper triangular matrix to produce an unbias filtered output of the first spatial data stream. Multiplying.

  Some embodiments of the present disclosure provide an apparatus for performing bias removal from a received signal obtained after minimum mean square error (MMSE) equalization in a wireless communication system. The apparatus generally includes logic for permuting an extension matrix of channel estimates of a plurality of spatial streams and a unitary matrix of the plurality of spatial streams to generate a permutation extension matrix of the channel estimates of the plurality of spatial streams. And a logic to perform QR decomposition of the permutation augmentation matrix to generate an upper triangular matrix and a first spatial to generate a bias filtered output of the first spatial stream of the spatial streams. Corresponding upper triangle to the bias filtered output to generate the logic for rotating the received signal of the stream using the corresponding unitary matrix and the unbias filtered output of the first spatial data stream Logic for multiplying the diagonal elements of the matrix.

  Some embodiments of the present disclosure provide an apparatus for performing bias removal from a received signal obtained after minimum mean square error (MMSE) equalization in a wireless communication system. The apparatus generally includes means for permuting an extension matrix of channel estimates of a plurality of spatial streams and a unitary matrix of the plurality of spatial streams to generate a permutation extension matrix of the channel estimates of the plurality of spatial streams. And means for performing QR decomposition of the permutation augmentation matrix to generate an upper triangular matrix, and a first spatial to generate a bias filtered output of the first spatial stream of the spatial streams Means for rotating the received signal of the stream using a corresponding unitary matrix and corresponding upper triangles to the bias filtered output to generate an unbias filtered output of the first spatial data stream Means for multiplying the diagonal elements of the matrix.

  Some embodiments of the present disclosure are generally obtained after minimum mean square error (MMSE) equalization in a wireless communication system comprising a computer-readable medium having instructions executable by one or more processors. A computer program product for performing bias removal from a received signal is included. This instruction generally includes instructions for replacing an extension matrix of channel estimates of a plurality of spatial streams and a unitary matrix of the plurality of spatial streams to generate a permutation extension matrix of the channel estimates of the plurality of spatial streams. And an instruction to perform QR decomposition of the permutation augmentation matrix to generate an upper triangular matrix and a first spatial to generate a bias filtered output of the first spatial stream of the spatial streams Instructions for rotating the received signal of the stream using the corresponding unitary matrix and the corresponding upper triangle to the bias filtered output to generate an unbias filtered output of the first spatial data stream Instructions for multiplying the diagonal elements of the matrix.

  For a better understanding of the above features of the present disclosure, a more specific description, briefly summarized above, may be obtained by reference to the embodiments, some of which are illustrated in the accompanying drawings. . However, the attached drawings illustrate only some exemplary embodiments of the present disclosure, and therefore the description should not be considered as limiting the scope, as it leads to other equally valid embodiments. Please note that.

1 illustrates an example wireless communication system, in accordance with some embodiments of the present disclosure. FIG. FIG. 3 illustrates various components that may be utilized in a wireless device according to some embodiments of the present disclosure. FIG. 3 illustrates an example transmitter and an example receiver that can be used within a wireless communication system in accordance with some embodiments of the present disclosure. 1 is a block diagram of a MIMO-OFDM wireless system used to derive a system model. FIG. 3 is a block diagram illustrating equivalence between MMSE equalization of the original channel and zero-forcing (ZF) equalization of the expanded channel. FIG. 3 is an exemplary block diagram of QRMMSE detection with bias removal applied in a MIMO-OFDM system with two transmit antennas and two receive antennas. FIG. 4 shows a process for QRMMSE detection with bias removal applied in a MIMO-OFDM system. The figure which shows the example component which can perform the calculation shown in FIG. Graph of bit error rate performance for conventional MMSE detection (biased and unbiased) and QRMMSE detection (biased and unbiased). FIG. 4 shows a process for QRMMSE-VBLAST detection with bias removal applied in a MIMO-OFDM system. The figure which shows the example component which can perform the calculation shown in FIG. FIG. 4 is an exemplary block diagram of QRMMSE-VBLAST detection with bias removal applied in a MIMO-OFDM system with two transmit antennas and two receive antennas.

  The word “exemplary” is used herein to mean “serving as an example, instance, or illustration”. Any embodiment described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments.

Exemplary Wireless Communication System The techniques described herein may be used for various communication systems including broadband wireless communication systems based on orthogonal multiplexing schemes. Examples of such communication systems include orthogonal frequency division multiple access (OFDMA) systems, single carrier frequency division multiple access (SC-FDMA) systems, and the like. An OFDMA system utilizes orthogonal frequency division multiplexing (OFDM), which is a modulation technique that partitions the overall system bandwidth into multiple orthogonal subcarriers. These subcarriers can also be called tones, bins, etc. In OFDM, each subcarrier can be independently modulated with data. SC-FDMA systems are interleaved FDMA (IFDMA) for transmitting on subcarriers distributed over the system bandwidth, local FDMA (LFDMA) for transmitting on adjacent subcarrier blocks, or adjacent Enhanced FDMA (EFDMA) for transmission on multiple blocks of subcarriers can be utilized. In general, modulation symbols are sent in the frequency domain with OFDM and in the time domain with SC-FDMA.

  One specific example of a communication system based on the orthogonal multiplexing scheme is a WiMAX system. WiMAX, representing Worldwide Interoperability for Microwave Access, is a standards-based broadband wireless technology that provides high-throughput broadband connectivity over long distances. Currently, there are two main applications of WiMAX: fixed WiMAX and mobile WiMAX. Fixed WiMAX applications are, for example, point-to-multipoint that enables broadband access in homes and businesses. Mobile WiMAX provides full mobility for cellular networks at broadband speeds.

  IEEE 802.16x is an emerging standards organization that defines an air interface for fixed and mobile broadband wireless access (BWA) systems. These standards define at least four different physical layers (PHYs) and one medium access control (MAC) layer. Of the four physical layers, the OFDM and OFDMA physical layers are most common in the fixed and mobile BWA regions, respectively.

  FIG. 1 illustrates an example of a wireless communication system 100 in which embodiments of the present disclosure can be employed. The wireless communication system 100 may be a broadband wireless communication system. The wireless communication system 100 can provide communication to a number of cells 102, each cell being served by a base station 104. Base station 104 may be a fixed station that communicates with user terminal 106. Base station 104 may alternatively be referred to as an access point, Node B, or some other terminology.

  FIG. 1 shows various user terminals 106 scattered throughout the system 100. The user terminal 106 may be fixed (that is, stationary) or moved. User terminal 106 may alternatively be referred to as a remote station, access terminal, terminal, subscriber unit, mobile station, station, user equipment, etc. User terminal 106 may be a wireless device such as a cellular phone, personal digital assistant (PDA), handheld device, wireless modem, laptop computer, personal computer.

  Various algorithms and methods may be used for transmission in the wireless communication system 100 between the base station 104 and the user terminal 106. For example, signals can be transmitted and received between base station 104 and user terminal 106 in accordance with OFDM / OFDMA techniques. In this case, the wireless communication system 100 can be referred to as an OFDM / OFDMA system.

  A communication link that enables transmission from the base station 104 to the user terminal 106 is referred to as a downlink (DL) 108, and a communication link that enables transmission from the user terminal 106 to the base station 104 is referred to as an uplink (UL) 110. Can be called. Alternatively, the downlink 108 can be referred to as a forward link or forward channel and the uplink 110 can be referred to as a reverse link or reverse channel.

  The cell 102 can be divided into a plurality of sectors 112. Sector 112 is a physical coverage area within cell 102. A base station 104 in the wireless communication system 100 may utilize an antenna that concentrates the flow of power in a particular sector 112 of the cell 102. Such an antenna can be called a directional antenna.

  FIG. 2 illustrates various components that may be utilized in a wireless device 202 that may be employed within the wireless communication system 100. The wireless device 202 is an example of a device that can be configured to implement the various methods described herein. Wireless device 202 may be base station 104 or user terminal 106.

  The wireless device 202 can include a processor 204 that controls the operation of the wireless device 202. The processor 204 is sometimes referred to as a central processing unit (CPU). Memory 206, which can include both read only memory (ROM) and random access memory (RAM), provides instructions and data to processor 204. A portion of memory 206 may also include non-volatile random access memory (NVRAM). The processor 204 generally performs logic and arithmetic operations based on program instructions stored in the memory 206. The instructions in memory 206 can be executed to implement the methods described herein.

  The wireless device 202 can also include a housing 208 that can include a transmitter 210 and a receiver 212 to allow transmission and reception of data between the wireless device 202 and a remote location. The transmitter 210 and the receiver 212 can be combined to form the transceiver 214. A plurality of transmit antennas 216 are attached to the housing 208 and are electrically coupled to the transceiver 214. Wireless device 202 may also include multiple transmitters, multiple receivers, and multiple transceivers (not shown).

  The wireless device 202 can also include a signal detector 218 that can be used to detect and quantify the level of the signal received by the transceiver 214. The signal detector 218 can detect signals such as total energy, energy per subcarrier per symbol, power spectral density, and other signals. The wireless device 202 may also include a digital signal processor (DSP) 220 for use in processing signals.

  The various components of the wireless device 202 can be coupled together by a bus system 222 that can include a power bus, a control signal bus, and a status signal bus in addition to a data bus.

  FIG. 3 shows an example of a transmitter 302 that can be used within a wireless communication system 100 that utilizes OFDM / OFDMA. The portion of transmitter 302 is implemented in transmitter 210 of wireless device 202. A transmitter 302 is implemented in the base station 104 to transmit data 306 to the user terminal 106 on the downlink 108. A transmitter 302 is also implemented in the user terminal 106 for transmitting data 306 over the uplink 110 to the base station 104.

  The transmitted data 306 is shown as being provided as an input to a serial to parallel (S / P) converter 308. The S / P converter 308 divides the transmission data into M parallel data streams 310.

  The M parallel data streams 310 are then provided as inputs to the mapper 312. Mapper 312 may map M parallel data streams 310 to M constellation points. The mapping is performed using some modulation constellation such as two phase shift keying (BPSK), four phase shift keying (QPSK), eight phase shift keying (8PSK), quadrature amplitude modulation (QAM). Accordingly, the mapper 312 outputs M parallel symbol streams 316, each corresponding to one of the M orthogonal subcarriers of the inverse fast Fourier transform (IFFT) 320. These M parallel symbol streams 316 are represented in the frequency domain and converted to M parallel time domain sample streams 318 by IFFT component 320.

Next, a brief note about the term is given. M parallel modulations in the frequency domain are equal to M modulation symbols in the frequency domain, which is equal to M mappings and M point IFFTs in the frequency domain, which is one (useful in the time domain). N) OFDM symbol, which is equal to M samples in the time domain. One OFDM symbol in the time domain, N _s, is equal to N _cp (number of guard samples per OFDM symbol) + M (number of useful samples per OFDM symbol).

  The M parallel time domain sample streams 318 are converted to OFDM / OFDMA symbol streams 322 by a parallel to serial (P / S) converter 324. The guard insertion component 326 inserts a guard interval between consecutive OFDM / OFDMA symbols in the OFDM / OFDMA symbol stream 322. The signal from guard insertion component 326 is then input to demultiplexer 340 to generate various data streams for multiple transmit antennas (or equivalently, spatial subchannels). Thereafter, the radio frequency (RF) front end 328 upconverts the baseband data stream for each antenna to the desired transmission frequency band, and then the antenna array 330 converts the resulting signal 332 into a plurality of spatial subchannels 334. Send over.

  FIG. 3 also illustrates an example of a receiver 304 that can be used within a wireless device 202 that utilizes OFDM / OFDMA. The portion of receiver 304 is implemented in receiver 212 of wireless device 202. Receiver 304 is implemented in user terminal 106 for receiving data 306 from base station 104 on downlink 108. Receiver 304 is also implemented in base station 104 to receive data 306 from user terminal 106 on uplink 110.

  The transmission signal 332 is shown moving on a plurality of spatial subchannels 334. When signal 332 'is received by antenna array 330', received signal 332 'is downconverted to a baseband signal by RF front end 328' and converted to a single stream by multiplexer 340 '. The guard removal component 326 'then removes the guard interval inserted between the OFDM / OFDMA symbols by the guard insertion component 326.

  The output of guard removal component 326 'is provided to S / P converter 324'. S / P converter 324 'splits OFDM / OFDMA symbol stream 322' into M parallel time-domain symbol streams 318 ', each corresponding to one of M orthogonal subcarriers. A Fast Fourier Transform (FFT) component 320 'converts the M parallel time domain symbol streams 318' to the frequency domain and outputs M parallel frequency domain symbol streams 316 '.

  The demapper 312 'performs the inverse of the symbol mapping operation performed by the mapper 312 and thereby outputs M parallel data streams 310'. P / S converter 308 'combines M parallel data streams 310' into a single data stream 306 '. Ideally, this data stream 306 ′ corresponds to the data 306 supplied as input to the transmitter 302. Note that elements 308 ', 310', 312 ', 316', 320 ', 318' and 324 'can all be found on baseband processor 350'.

Exemplary MIMO-OFDM System Model FIG. 4 shows a block diagram of a typical multiple-input multiple-output (MIMO) OFDM wireless communication system with N _t transmit antennas and N _r receive antennas. The system model for the kth subcarrier (frequency subchannel) can be expressed by the following linear equation:

_Where N _fft is the number of orthogonal subcarriers (frequency bins) in the MIMO wireless system.

In the following equations and the accompanying disclosure, the subcarrier index k is omitted for simplicity. Therefore, the system model can be rewritten with the simple notation as follows.

Where y is the [N _r × 1] received symbol vector, H is the [N _r × N _t ] channel matrix, and h _j is between transmit antenna j and all N _r receive antennas. including channel gain, a j-th column vector of H, x is [n _t × 1] is the transmitted symbol vector, n represents with covariance matrix _{^{E (nn H) [n r}} × 1] complex noise Is a vector.

As shown in FIG. 4, the transmission signal is first encoded by the MIMO encoder 410. Redundancy can be included to protect information data during transmission over noisy wireless channels. Then, as shown in FIG. 4, the encoded signal is divided into N _t spatial data streams x ₁ , x ₂ ,. . . , X _Nt . Inverse Fast Fourier Transform (IFFT) units 412 ₁ ,. . . . By using 412 _Nt , multiple spatial data streams can be converted to the time domain. Then, the signal is up-converted to a desired transmission frequency band, N _r · N through the _t pieces of spatial subchannels N _t transmit antennas 414 _1,. . . 414 _Nt .

At the receiver, N _r receive antennas 416 ₁ ,. . . . 416 _Nr is employed. Fast Fourier transform (FFT) units 418 ₁ ,. . . . By using 418 _Nr , the received data stream can be transformed back into the frequency domain. The frequency domain signal may be input to a MIMO detector 420 that generates a reliability message for encoded bits transmitted over multiple spatial subchannels. The reliability message represents the probability that a particular transmission coded bit is bit “0” or bit “1”. This information can be passed to the outer MIMO channel decoder 422 and estimated information data for multiple spatial subchannels (transmit antennas).

Becomes available after removing the redundancy included in the transmitter.

Exemplary Linear MMSE Detection For a well-known linear MMSE detector (channel equalizer), G _MMSE can be expressed in the form:

Where

Is the variance of the noise at the receiver,

Is a unit matrix of size [N _t × N _t ]. The signal obtained after applying MMSE filtering can be expressed as:

A linear system model described by equation (2) is shown in FIG. Transmission of signal x over multiple spatial subchannels using a particular frequency subchannel is indicated by block 510 and the effect of channel noise at the receiver is indicated by summer 512. Then, as indicated by block 520, MMSE filtering is applied to the received symbol vector y.

The error covariance matrix for MMSE channel equalization is equal to:

Expanded channel, also shown in FIG.

A system model with can be defined as follows.

Where

Is the expanded channel matrix (unit 530);

Is the expanded received signal vector,

Is the expanded complex noise vector added to the received vector (indicated by adder 532);

Is [(N _r + N _t ) × N _t ],

Is [(N _r + N _t ) × 1], x is [N _t × 1],

Is [(N _r + N _t ) × 1].

Expansion channel

It can be shown that the zero forcing (ZF) equalizer is equivalent to the original channel H MMSE equalization. The ZF equalizer for the expanded channel can be defined as follows:

When replacing an enlarged portion y _a in 0 (unit 540), i.e.,

The following equation holds after ZF equalization (unit 550).

After comparing equations (13) and (8), the expanded channel matrix

It can be concluded that the ZF equalization is equivalent to the MMSE equalization of the original matrix H. Figure 5 shows the equivalence of these two approaches, ie the filtered output from equation (8).

And equation (13) can be the same.

The rotation of the received signal by the unitary matrix Q ^H (Hermitian version of the unitary matrix obtained from the QR decomposition of the original channel matrix) results in the same signal that is generated using ZF equalization. Thus, the equivalence of the algorithm shown in FIG. 5 may lead to the ability to implement MMSE detection using QR decomposition of the expanded channel matrix.

Exemplary Unbiased MMSE Detection Using QR Decomposition To illustrate MMSE detection based on QR decomposition (QRMMMSE detection), a wireless system with two transmit antennas can be described without loss of generality. An exemplary block diagram of a MIMO receiver that performs QRMMSE detection of two spatial data streams transmitted over two spatial subchannels (a pair of transmit antennas) is shown in FIG.

  First the channel matrix is expanded (unit 610), and then the columns are replaced so that the last column of the resulting permutation matrix corresponds to the spatial subchannel (spatial data stream) currently being decoded. . In the illustrated example, the permutation units 620 and 650 in FIG. 6 are applied for decoding the first and second spatial data streams, respectively. QR decomposition of the permutation channel matrix can be performed by units 630 and 660 for decoding the first and second spatial data streams, respectively.

The QR decomposition of the expanded channel matrix can be expressed as:

Where the matrix

Consists of orthonormal column vectors,

Is an upper triangular matrix with real diagonal elements. From equation (14), it is as follows.

In the above equation, Q ₁ is [N _r × N _t ] and Q ₂ is [N _t × N _t ].

As shown in equation (13), instead of performing MMSE equalization of the original channel matrix, a zero forcing (ZF) operation can be applied to the expanded channel matrix. Units 640 and 670 in FIG. 6 perform zero forcing filtering (rotation) of the received vector y, and the result is the bias filtered output of the first and second spatial data streams, respectively. From equation (12), it can be seen that it may be necessary to rotate the received symbol vector y only for the upper half of the unitary matrix Q (matrix Q _{1 in} equation (14)). Thus, the filtered output with the bias introduced by the MMSE operation can be expressed as:

From equation (17), it is as follows.

For an exemplary MIMO-OFDM system with two transmit antennas and two receive antennas, the following equation can be specified:

As an illustrative example, decoding of the _second spatial data stream x ₂ can be considered. According to equation (15), the matrix Q ₂ is an upper triangle. Therefore, equation (19) can be rewritten as follows.

Term

Represents the bias for the signal transmitted from the first antenna (first spatial data stream) and

Represents the bias for the signal transmitted from the second antenna (second spatial data stream).

As an example, consider first removing the bias from the second spatial data stream. In one embodiment, the second bias data stream is a bias term.

The bias can be removed by dividing by. However, division is not a desirable arithmetic operation for efficient hardware implementation due to the long latency and the large area size of the divider. Division operations may also generate very complex noise interference terms that prevent accurate calculation of the noise variance required for calculating the log likelihood ratio (LLR) of the coded bits. Instead of conventional bias removal techniques based on direct division operations, this disclosure proposes a simpler bias removal technique.

The last diagonal element r of the upper triangular matrix R (shown by multipliers 676 and 646 in FIG. 6 for decoding the second and first spatial streams, respectively) in the second row of equation (21) _When multiplying by ₂₂ , the unbiased filtered output of the second spatial stream can be obtained from the following equation:

From equation (15), r ₂₂ q ₄ = σ _n (element q ₄ is a real number), so the effective channel of the second spatial data stream can be obtained from the following equation:

The calculation of the effective subchannel corresponding to the second spatial stream is shown in FIG. 6 by applying a square multiplier 672 and applying a subtraction unit 674 to the last diagonal element r ₂₂ of the matrix R. Has been. A similar processing flow can be applied for the calculation of the effective channel corresponding to the first spatial data stream, which is represented by a square multiplier 642 and a subtraction unit 644.

Regarding noise dispersion,

Since the transmission signal x ₂ is independent of the noise components related to the first and second spatial subchannels, the effective noise variance corresponding to the second spatial subchannel can be calculated as follows.

From equation (14)

It can be seen that it is. Then, equation (24) becomes as follows.

After the bias is removed, the effective signal model for the second spatial subchannel is:

The log likelihood ratio (LLR) of the coded bits corresponding to the second spatial subchannel, as indicated by unit 680 in FIG.

When,

When,

And can be calculated by using The signal model for the calculation of the LLR of the second data stream is as follows:

Where

,and

It is. The calculation of the LLRs of the coded bits belonging to the first spatial data stream can be performed in a similar manner, which is also shown in FIG.

When

Are input to the LLR calculation unit 680.

When prewhitening is applied to the channel estimates and received samples, the effective noise variance at the receiver is unitary.

There is no need to calculate. Next, the signal model for the calculation of the LLR corresponding to the second spatial data stream is as follows.

Where n _{2 to} N (0,1) and

It is. From FIG. 6, it can be seen that the same implementation logic can be used for the detection of any spatial data stream with appropriate permutation of the channel matrix.

The process of decoding a spatial data stream using QRMMSE detection can be generalized for a system with N _t transmit antennas (spatial subchannels), as shown by the flow diagram in FIG. In 712, first, the [N _r × N _t ] channel matrix H is expanded as in equation (14), and the [(N _r × N _t ) × N _t ] channel matrix is expanded.

Get. Subsequently, at 714, every spatial data stream j = 1, 2,. . . , N _t , permutation matrix

Of the expanded channel matrix such that the rightmost column of corresponds to the j th decoded stream

Can be substituted.

Then, at 716, every spatial data stream j = 1, 2,. . . , N _t

So that the QR decomposition of the permutation expanded channel matrix defined by equation (29) can be performed. Thereafter, every spatial data stream j = 1, 2,. . . , N _t , the permutation expansion matrix

The corresponding Hermitian version of the unitary matrix Q ₁ (matrix), which is the upper half of the unitary matrix obtained from the QR decomposition of

) To the received signal vector y, the received signal vector y can be rotated. As a result of this rotation, at 718, a bias filtered output of any spatial data stream can be obtained according to equation (18).

720, every spatial data stream j = 1, 2,. . . , N _t , the biased filtered output of the j th data stream

To obtain the last diagonal element of the upper triangular matrix R as the last element of the bias filtered output

Can be multiplied. Following this, at 722, every spatial data stream j = 1, 2,. . . , N _t , the effective channel and effective noise variance can be obtained as follows.

Thus, the non-bias filtered output of the jth spatial data stream can be expressed as:

Next, the signal model for the jth spatial subchannel is:

At 724, according to this signal model,

And h _eff

And each spatial data stream j = 1, 2,. . . , N _t encoded bits LLR can be calculated. The calculated LLR can be passed to the outer channel decoder for estimation of all N _t spatial data streams.

  The unbiased QRMMSE algorithm is equivalent to the conventional unbiased MMSE approach with respect to detection accuracy. FIG. 8 shows a graph of bit error rate (BER) performance of conventional biased and unbiased MMSE equalization versus biased and unbiased QRMMSE equalization. A simulation is performed for a wireless system with two transmit antennas and two receive antennas, 16QAM modulation is applied at the transmitter, a rate 1/2 convolutional code is used, and a channel D fading model is assumed.

  Curve 810 in FIG. 8 represents BER performance for conventional biased MMSE detection, curve 820 represents BER performance for biased QRMMSE detection, and curve 830 represents BER performance for conventional unbiased MMSE detection. Curve 840 represents the BER performance for QRMMSE detection using the proposed bias removal technique. It can be seen that the same BER is obtained for the conventional bias MMSE and bias QRMMSE detection algorithms. If the bias term is removed from the QRMMSE filtered output based on the technique proposed in this disclosure, a substantial BER performance improvement can be achieved (curve 840 vs. curve 820 in FIG. 8).

  The advanced MIMO receiver can comprise QR decomposition logic for multi-stream detection. When the user equipment is on the boundary of two cells (eg, during the soft handover process), multi-stream detection may not be as effective as when a simple MMSE operation is in a better working cell. In this case, the receiver can switch to MMSE detection to improve error rate performance. By applying the proposed bias removal technique, a QRMMSE receiver can be efficiently implemented by using QR decomposition logic.

  As presented herein, one function of the proposed unbiased QRMMSE equalization is a simple bias removal technique. This approach is less complex than conventional bias removal algorithms based on division operations and can achieve substantial improvements in error rate performance.

Exemplary QRMMSE-VBLAST Receiver The well-known Vertical-Bell Laboratories-Layered-Space-Time (VBLAST) approach can be combined with QRMMSE detection (hereinafter abbreviated as VBLAST-QRMMMSE). The principle of the VBLAST algorithm is to first decode the most reliable signal of the multiple spatial data streams, generally the signal with the highest signal to noise ratio (SNR). The decoded signal can then be subtracted (deleted) from all bias space signals obtained after QRMMSE, thus generating a new signal with reduced multiple access interference. The decoding and deletion procedure can be repeated for all utilized spatial subchannels in the communication system.

FIG. 9 shows a process for estimating N _t spatial data streams by utilizing VBLAST based QRMMSE detection. At 910, the aforementioned unbiased QRMMSE detection is performed for all utilized spatial data streams j = 1,. . . , N _t . Thereafter, at 912, the spatial data stream j ^* defined by equation (29) and its corresponding permutation channel matrix are used to determine that this particular spatial stream has the best error rate performance of all N _t spatial streams. Choose to give. The selected spatial data stream j ^* generally corresponds to the spatial subchannel with the highest SNR. The permutation channel matrix corresponding to the selected spatial data stream j ^* may be utilized for further processing. At 914, a spatial data stream

And remove the scaled decoded component from all the bias spatial data streams obtained based on the bias QRMMSE. In some embodiments of the present disclosure, it may be assumed that j ^* = N _t without loss of generality.

For best performance, the remaining N _t −1 spatial data streams need to be decoded in order from the most reliable spatial stream to the lowest spatial stream. This entails repeating the selection procedure from 912, including searching for the best permutation matrix of all spatial streams that have not yet been decoded. Instead, the matrix found at 912 can be used until all data streams have been decoded. This method is shown in FIG. Since the N _t th data stream has already been decoded at 912 and subtracted at 914, the data stream index j starts at N _t −1. At 918, the bias signal corresponding to the currently decoded spatial data stream j is multiplied by the jth diagonal element r _jj of the upper triangular matrix R for bias removal. The upper triangular matrix R is obtained as a result of the QR decomposition of the permutation channel matrix selected at 912. Following bias removal, it decodes the N _t th data stream at 914, just like as deleted, then decodes the j-th spatial data streams, and then subtracted from all remaining bias spatial data stream (deleted) To do. This decoding process is repeated N _t −1 times corresponding to the number of remaining spatial data streams to be decoded (steps 918-924 in FIG. 9 are repeated N _t −1 times).

  As an illustrative example, the QRMMSE-VBLAST detection algorithm can be utilized for a MIMO-OFDM system with two transmit antennas (decoding two spatial data streams). An exemplary block diagram of such a QRMMSE-VBLAST receiver for detection of two spatial data streams is shown in FIG.

  QRMMSE operations can be applied by unit 1010 to generate bias signals for both spatial subchannels. In this exemplary case, it can be assumed that the second spatial data stream is more reliable and is therefore decoded first.

In unit 1012, a bias filtered output, by multiplying the last diagonal element r ₂₂ of the upper triangular matrix R generated after QR decomposition of expanding channel matrix from equation (14), a second data stream

To obtain an unbiased filtered output. Unbiased filtered output

And an effective channel corresponding to the second spatial stream

And effective noise variance corresponding to the second spatial subchannel

To obtain the LLR of the coded bits corresponding to the second spatial data stream. The estimated second spatial data stream is then used by the channel decoder 1020 using the LLR.

Get.

In multiplier 1018, if the estimated signal is scaled by element r ₁₂ , the element corresponding to the first row and second column of matrix R, the bias in the first row of equation (21) A second spatial data stream estimated from the signal

Is deleted. Then, from the first row of equation (21), the scaled estimated signal

Is subtracted (addition / subtraction unit 1014). The remaining signal represents the bias estimate of the first spatial data stream and is equal to:

When multiplying equation (35) by the first diagonal element r ₁₁ of matrix R (unit 1016), the following equation holds:

From equation (15), r ₁₁ p ₃ = σ _n is true. Then, an effective channel corresponding to the first spatial stream is obtained as follows.

For the effective noise variance corresponding to the first spatial subchannel, from equation (14)

Therefore, the following can be derived.

Thus, the unbiased filtered output of the first spatial data stream can be expressed as:

, H _eff , and n _eff are used to calculate the LLR of the coded bits transmitted from the first spatial subchannel (transmitted from the first antenna), and decoder unit 1030 The first spatial data stream decoded by

Get.

  By applying an unbiased QRMMSE-VBLAST approach, multiple access interference can be further reduced compared to unbiased QRMMSE detection. On the other hand, the continuous nature of the algorithm may result in additional processing latency.

  Various operations of the methods described above may be performed by various hardware (s) and / or software components and / or modules corresponding to the means plus function block shown in the figure. In general, if the method is illustrated in a diagram having a corresponding relative means plus function diagram, the operational block corresponds to a means plus function block having a similar numbering. For example, blocks 712-724 shown in FIG. 7 correspond to means-plus-function blocks 712A-724A shown in FIG. 7A. Similarly, blocks 912 to 922 shown in FIG. 9 correspond to means plus function blocks 912A to 922A shown in FIG. 9A.

  Various exemplary logic blocks, modules, and circuits described in connection with this disclosure may be general purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate array signals (FPGAs), or It can be implemented or performed using other programmable logic devices (PLDs), individual gate or transistor logic, individual hardware components, or any combination thereof designed to perform the functions described herein. A general purpose processor may be a microprocessor, but in the alternative, the processor may be any commercially available processor, controller, microcontroller or state machine. A processor may be implemented as a combination of computing devices, eg, a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors associated with a DSP core, or any other such configuration. You can also.

  The method or algorithm steps described in connection with the present disclosure may be implemented in direct hardware, software modules executed by a processor, or a combination of the two. A software module may reside in any form of storage medium that is known in the art. Some examples of storage media that can be used include random access memory (RAM), read only memory (ROM), flash memory, EPROM memory, EEPROM memory, registers, hard disk, removable disk, CD-ROM, and the like. A software module can comprise a single instruction, or multiple instructions, and can be distributed over several different code segments, between different programs, and across multiple storage media. A storage medium may be coupled to the processor such that the processor can read information from, and write information to, the storage medium. In the alternative, the storage medium may be integral to the processor.

  The methods disclosed herein comprise one or more steps or actions for achieving the described method. The method steps and / or actions may be interchanged with one another without departing from the scope of the claims. In other words, unless a specific order of steps or actions is specified, the order and / or use of specific steps and / or actions may be changed without departing from the scope of the claims.

  The described functionality can be implemented in hardware, software, firmware, or a combination thereof. If implemented in software, the functions can be stored on a computer-readable medium as one or more instructions. A storage media may be any available media that can be accessed by a computer. By way of example, and not limitation, such computer readable media can be RAM, ROM, EEPROM, CD-ROM, or other optical disk storage, magnetic disk storage or other magnetic storage device, or any desired form in the form of instructions or data structures. Any other medium that can be used to carry or store the program code and that can be accessed by a computer can be provided. Discs and discs used in this specification are compact discs (CD), laser discs, optical discs, digital versatile discs (DVDs), floppy discs (discs). (Registered trademark) disk, and Blu-ray (registered trademark) disc, which normally reproduces data magnetically, and the disc optically reproduces data with a laser. To play.

  Software or instructions can also be transmitted over a transmission medium. For example, the software can use a coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technology such as infrared, wireless, and microwave, from a website, server, or other remote source When transmitted, coaxial technologies, fiber optic cables, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of transmission media.

  Moreover, modules and / or other suitable means for performing the methods and techniques described herein can be downloaded and / or otherwise obtained by user terminals and / or base stations when applicable. I want to be understood. For example, such a device can be coupled to a server to allow transfer of means for performing the methods described herein. Alternatively, the various methods described herein are storage means (e.g., such that the user terminal and / or base station can obtain various methods when the storage means is coupled or provided to the device). , RAM, ROM, a physical storage medium such as a compact disk (CD) or a floppy disk). Further, any other suitable technique for providing the devices with the methods and techniques described herein may be utilized.

  It is to be understood that the claims are not limited to the precise configuration and components illustrated above. Various modifications, changes and variations may be made in the arrangement, operation and details of the methods and apparatus described above without departing from the scope of the claims.

Claims (16)

A method for performing bias removal from a received signal obtained after minimum mean square error (MMSE) equalization in a wireless communication system, comprising:
Append a matrix of channel gain estimates with a scaled identity matrix to generate an expanded matrix of channel estimates,
Replacing an expansion matrix of channel estimates of the plurality of spatial streams to generate a replacement expansion matrix of channel estimates of the plurality of spatial streams;
Performing QR decomposition of the permutation augmentation matrix to generate a unitary matrix and an upper triangular matrix of the plurality of spatial streams;
Rotating the received signal of the first spatial stream with a corresponding unitary matrix to generate a bias filtered output of the first spatial stream of the spatial streams;
Multiplying the bias filtered output by a diagonal element of the corresponding upper triangular matrix to generate an unbias filtered output of the first spatial data stream. Calculating effective channel estimates for multiple spatial subchannels;
Calculating an effective noise variance of the plurality of spatial subchannels;
Calculating a log-likelihood ratio of coded bits transmitted over the spatial subchannel by using an unbiased filtered output, an effective channel estimate, and an effective noise variance. The method of claim 1. Removing the decoded unbiased output of the first spatial stream from the bias signal of the remaining spatial stream;
Continuously repeating decoding a plurality of spatial streams based on bias removal of a plurality of spatial subchannels and removing the decoded unbiased output from the remaining bias spatial streams;
2. The method of claim 1, further comprising multiplying a bias received signal of the remaining spatial subchannel by a diagonal element of the upper triangular matrix to generate a biased filtered output of the remaining spatial subchannel. The method described. Calculating effective estimates for multiple spatial subchannels;
Calculating the effective noise variance of multiple spatial subchannels;
Compute the log-likelihood ratio of coded bits transmitted over multiple spatial subchannels by using the corresponding unbiased filtered output, effective channel estimate, and effective noise variance The method of claim 3, further comprising: An apparatus for performing bias removal from a received signal obtained after minimum mean square error (MMSE) equalization in a wireless communication system, comprising:
Logic to append a matrix of channel gain estimates with a scaled identity matrix to generate an expanded matrix of channel estimates;
Logic to replace an expansion matrix of channel estimates of the plurality of spatial streams to generate a replacement extension matrix of channel estimates of the plurality of spatial streams;
Logic to perform QR decomposition of the permutation augmentation matrix to generate unitary and upper triangular matrices of the plurality of spatial streams;
Logic for rotating the received signal of the first spatial stream using a corresponding unitary matrix to generate a bias filtered output of the first spatial stream of the spatial streams;
An apparatus comprising: logic for multiplying the bias filtered output by a corresponding upper triangular matrix diagonal element to produce an unbias filtered output of the first spatial data stream. Logic for calculating effective channel estimates for multiple spatial subchannels;
Logic for calculating an effective noise variance of the plurality of spatial subchannels;
Logic for calculating the log-likelihood ratio of the coded bits transmitted over the spatial subchannel by using the unbiased filtered output, the effective channel estimate, and the effective noise variance; The apparatus according to claim 5, further comprising: Logic for removing the decoded unbiased output of the first spatial stream from the bias signal of the remaining spatial stream;
Logic to continuously iterate decoding a plurality of spatial streams based on bias removal of a plurality of spatial subchannels and removing the decoded unbiased output from the remaining bias spatial streams;
And further comprising logic for multiplying the biased received signal of the remaining spatial subchannel by a diagonal element of the upper triangular matrix to generate a biased filtered output of the remaining spatial subchannel. 5. The apparatus according to 5. Logic for calculating effective estimates of multiple spatial subchannels;
Logic for calculating the effective noise variance of multiple spatial subchannels;
To calculate the log-likelihood ratio of coded bits transmitted over multiple spatial subchannels by using the corresponding unbiased filtered output, effective channel estimate, and effective noise variance The apparatus of claim 7, further comprising: An apparatus for performing bias removal from a received signal obtained after minimum mean square error (MMSE) equalization in a wireless communication system, comprising:
Means for adding a matrix of channel gain estimates having a scaled identity matrix to generate an expanded matrix of channel estimates;
Means for replacing an expansion matrix of channel estimates of the plurality of spatial streams to generate a replacement extension matrix of channel estimates of the plurality of spatial streams;
Means for performing QR decomposition of the permutation augmentation matrix to generate unitary and upper triangular matrices of the plurality of spatial streams;
Means for rotating the received signal of the first spatial stream using a corresponding unitary matrix to generate a bias filtered output of the first spatial stream of the spatial streams;
Means for multiplying the bias filtered output by a diagonal element of a corresponding upper triangular matrix to generate an unbias filtered output of the first spatial data stream. Means for calculating an effective channel estimate for a plurality of spatial subchannels;
Means for calculating an effective noise variance of the plurality of spatial subchannels;
Means for calculating a log-likelihood ratio of coded bits transmitted over the spatial subchannel by using an unbiased filtered output, an effective channel estimate, and an effective noise variance; The apparatus of claim 9, further comprising: Means for removing the decoded unbiased output of the first spatial stream from a bias signal of the remaining spatial stream;
Means for continuously repeating decoding a plurality of spatial streams based on bias removal of a plurality of spatial subchannels and removing the decoded unbiased output from the remaining bias spatial streams;
Means for multiplying the remaining spatial subchannel bias received signal by a diagonal element of the upper triangular matrix to generate an unbiased filtered output of the remaining spatial subchannel. 9. The apparatus according to 9. Means for calculating an effective estimate of a plurality of spatial subchannels;
Means for calculating the effective noise variance of a plurality of spatial subchannels;
To calculate the log-likelihood ratio of coded bits transmitted over multiple spatial subchannels by using the corresponding unbiased filtered output, effective channel estimate, and effective noise variance The apparatus of claim 11, further comprising: For performing bias removal from a received signal obtained after minimum mean square error (MMSE) equalization in a wireless communication system comprising a computer readable medium having instructions executable by one or more processors A computer- readable recording medium , wherein the instructions are
Instructions for appending a matrix of channel gain estimates having a scaled identity matrix to generate an expanded matrix of channel estimates;
Instructions for replacing an expansion matrix of channel estimates of the plurality of spatial streams to generate a replacement extension matrix of channel estimates of the plurality of spatial streams;
Instructions for performing QR decomposition of the permutation augmentation matrix to generate unitary and upper triangular matrices of the plurality of spatial streams;
Instructions for rotating the received signal of the first spatial stream using a corresponding unitary matrix to generate a bias filtered output of the first spatial stream of the spatial streams;
Computer- readable comprising: instructions for multiplying the biased filtered output by a diagonal element of a corresponding upper triangular matrix to generate a biased filtered output of the first spatial data stream Recording medium . The instruction is
Instructions for calculating effective channel estimates for multiple spatial subchannels;
Instructions for calculating an effective noise variance of the plurality of spatial subchannels;
Instructions for calculating a log-likelihood ratio of coded bits transmitted over the spatial subchannel by using an unbiased filtered output, an effective channel estimate, and an effective noise variance; The computer- readable recording medium according to claim 13, further comprising: The instruction is
Instructions for removing the decoded unbiased output of the first spatial stream from the bias signal of the remaining spatial stream;
Instructions for continuously repeating decoding a plurality of spatial streams based on bias removal of a plurality of spatial subchannels and removing the decoded unbiased output from the remaining bias spatial streams;
Instructions for multiplying the biased received signal of the remaining spatial subchannel by a diagonal element of the upper triangular matrix to generate a biased filtered output of the remaining spatial subchannel. 14. A computer- readable recording medium according to 13. The instruction is
Instructions for calculating effective estimates of multiple spatial subchannels;
Instructions for calculating the effective noise variance of multiple spatial subchannels;
To calculate the log-likelihood ratio of coded bits transmitted over multiple spatial subchannels by using the corresponding unbiased filtered output, effective channel estimate, and effective noise variance The computer- readable recording medium according to claim 15, further comprising:

Images